In [2]:
#load the necessary packages:
import numpy as np
import pandas as pd
import cimcb as cb
from sklearn.model_selection import train_test_split
import seaborn as sns
np.random.seed(42)
print('All packages successfully loaded')

Using Theano backend.

All packages successfully loaded
In [3]:
# load the necessary file:
file = 'MTBLS136.xlsx'  

DataTable,PeakTable = cb.utils.load_dataXL(file, DataSheet='Data', PeakSheet='Peak') 

Loadings PeakFile: Peak
Loadings DataFile: Data
Data Table & Peak Table is suitable.
TOTAL SAMPLES: 1649 TOTAL PEAKS: 949
Done!
In [4]:
DataTable
Out[4]:
Idx SampleID Class Hormone M1 M2 M3 M4 M5 M6 ... M940 M941 M942 M943 M944 M945 M946 M947 M948 M949
1 1 063942_01 2.0 Nonuser NaN NaN NaN 95695.0 311381856 4701657.0 ... NaN 641183.0 NaN 29736404.0 146489552.0 25615702.0 125249.0 209356.0 1406015.0 8034717.0
2 2 037523_02 0.0 E-only NaN NaN NaN 503202.0 98715432 753959.0 ... 2544193.0 548504.0 NaN NaN NaN 39473920.0 192490.0 156234.0 808012.0 1031587.0
3 3 073016_03 0.0 E-only NaN NaN NaN 104038.0 145110144 NaN ... 403755.0 768151.0 NaN NaN NaN 25694854.0 233899.0 129411.0 1166341.0 630163.0
4 4 061210_01 2.0 Nonuser NaN NaN NaN 366910.0 156989408 549563.0 ... 561696.0 1008375.0 NaN NaN NaN 26806070.0 99558.0 NaN 2180112.0 1008026.0
5 5 063862_01 0.0 E-only NaN NaN NaN 608979.0 140037936 1326962.0 ... 48161.0 750182.0 NaN NaN NaN 33121192.0 NaN 125668.0 552264.0 1529796.0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
1645 1645 IU_4945_1001 NaN NaN NaN NaN NaN 115469.0 147413616 576361.0 ... 144390.0 338859.0 NaN NaN NaN 24838220.0 99708.0 64767.0 1244962.0 1651202.0
1646 1646 IU_5881_1000 NaN NaN NaN NaN NaN NaN 141172224 NaN ... NaN 511543.0 NaN NaN NaN 57732092.0 160562.0 NaN 1373740.0 475267.0
1647 1647 IU_5881_1001 NaN NaN NaN NaN NaN NaN 145479904 221693.0 ... NaN 563088.0 NaN NaN NaN 68145600.0 172044.0 57869.0 1522911.0 460609.0
1648 1648 IU_4945_1000 NaN NaN NaN NaN NaN 478657.0 156384400 731069.0 ... 166907.0 418370.0 NaN NaN NaN 26290880.0 116872.0 NaN 1116896.0 1980565.0
1649 1649 IU_5193_1000 NaN NaN NaN NaN NaN NaN 148171072 484084.0 ... 41187.0 392013.0 NaN NaN NaN 26832696.0 85120.0 71943.0 1281148.0 1059509.0

1649 rows × 953 columns

In [5]:
#the Idx column is dropped and SampleID column is made index.
DataTable= DataTable.drop(["Idx"], axis= 1)
DataTable= DataTable.set_index("SampleID")
In [9]:
# heatmap for observing some missing values.
cols = DataTable.columns
colors = ['#008000', '#ff0000']
sns.heatmap(DataTable[cols].isnull(), cmap=sns.color_palette(['#1f77b4', '#ff0000']),
            cbar=False, linecolor='gray')
Out[9]:
<AxesSubplot:ylabel='SampleID'>
In [8]:
# see if any NA in the whole df:
DataTable.isna().sum().sum()
Out[8]:
310432

Make the data dirty:¶

In [10]:
# randomly remove 10% of the values:
mask_10 = np.random.choice([True,False],size=DataTable.shape,p=[0.1,0.9])
miss_10_DataTable = DataTable.mask(mask_10,other=np.nan)
miss_10_DataSet = miss_10_DataTable.copy()# used later on for removing irrelevant rows
In [11]:
# randomly remove 50% of the values
mask_50 = np.random.choice([True,False],size=DataTable.shape,p=[0.5,0.5])
miss_50_DataTable = DataTable.mask(mask_50,other=np.nan)
miss_50_DataSet = miss_50_DataTable.copy()# used later on for removing irrelevant rows
In [12]:
# randomly change 10% of the values
change_10 = np.random.randint(low= 500000, high= 8000000,size= DataTable.shape)
change_10_DataTable = DataTable.mask(mask_10,other=change_10)
change_10_DataSet = change_10_DataTable.copy()# used later on for removing irrelevant data
In [13]:
# randomly change 50% of the values
change_50 = np.random.randint(low= 500000, high= 8000000,size= DataTable.shape)
change_50_DataTable = DataTable.mask(mask_50,other=change_50)
change_50_DataSet = change_50_DataTable.copy()# used later on for removing irrelevant data
In [12]:
# heatmap for observing some missing values.
cols = DataTable.columns
colors = ['#008000', '#ff0000']
sns.heatmap(DataTable[cols].isnull(), cmap=sns.color_palette(['#1f77b4', '#ff0000']),
            cbar=False, linecolor='gray')
Out[12]:
<AxesSubplot:ylabel='SampleID'>
In [13]:
# heatmap for observing some missing values.
cols = change_50_DataSet.columns
colors = ['#008000', '#ff0000']
sns.heatmap(change_50_DataSet[cols].isnull(), cmap=sns.color_palette(['#1f77b4', '#ff0000']),
            cbar=False, linecolor='gray')
Out[13]:
<AxesSubplot:ylabel='SampleID'>

cleaning the data:¶

Median imputation:¶

for 10% missing:¶

In [14]:
# impute the missing values and create the missing value indicator variables for each numeric column.
df_numeric = miss_10_DataTable._get_numeric_data().copy()
numeric_cols = df_numeric.columns.values
#impute using median for numberic columns:
for col in numeric_cols:
    df_numeric[col].fillna(df_numeric[col].median(), inplace=True)

#impute using mode() for non numeric columns
df_non_numeric = miss_10_DataTable.select_dtypes(exclude=np.number).copy()
non_numeric_cols = df_non_numeric.columns.values

for col in non_numeric_cols:
    df_non_numeric[col].fillna(df_non_numeric[col].mode()[0], inplace=True)
    
#concatinate both imputed non numeric and numeric data:
imputed_10_miss = pd.concat([df_non_numeric,df_numeric], axis=1, join='inner')
#plot the missing vs no missing after imputation:
cols = imputed_10_miss.columns
colors = ['#008000', '#ff0000']
sns.heatmap(imputed_10_miss[cols].isnull(), cmap=sns.color_palette(['#1f77b4', '#ff0000']),
            cbar=False, linecolor='gray')
Out[14]:
<AxesSubplot:ylabel='SampleID'>
In [15]:
# check if any missings remain in the data:
imputed_10_miss.columns[imputed_10_miss.isna().any()].tolist()
Out[15]:
[]

ML: SVM-LIN¶

In [16]:
# Clean PeakTable and Extract PeakList
PercMiss = PeakTable['Perc_missing']  
PeakTableClean = PeakTable[(PercMiss < 20)] 
PeakList = PeakTableClean['Name']  

# Select Subset of Data
DataTable2 = imputed_10_miss[(imputed_10_miss.Class == 1) | (imputed_10_miss.Class == 0)]

# Create a Binary Y Vector 
Outcomes = DataTable2['Class']
Y = Outcomes.values 

# Split Data into Train (2/3) and Test (1/3)
DataTrain, DataTest, YTrain, YTest = train_test_split(DataTable2, Y, test_size=1/3, stratify=Y, random_state=85)

# Extract Train Data 
XTrain = DataTrain[PeakList]                                    
XTrainLog = np.log(XTrain)                                          
XTrainScale, mu, sigma = cb.utils.scale(XTrainLog, method='auto', return_mu_sigma=True)              
XTrainKnn = XTrainScale# cb.utils.knnimpute(XTrainScale, k=3)    

# Extract Test Data
XTest = DataTest[PeakList]                                     
XTestLog = np.log(XTest)                                          
XTestScale = cb.utils.scale(XTestLog, method='auto', mu=mu, sigma=sigma) 
XTestKnn = XTestScale #cb.utils.knnimpute(XTestScale, k=3)
In [17]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...

1/2: 100%|███████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 304.37it/s]
2/2: 100%|██████████████████████████████████████████████████████████████████████████████| 70/70 [00:00<00:00, 71.17it/s]

Time taken: 1.29 minutes with 30 cores
Done!
Loading BokehJS ...
Loading BokehJS ...
In [18]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...

1/2: 100%|█████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 11550.01it/s]
2/2: 100%|██████████████████████████████████████████████████████████████████████████████| 70/70 [00:01<00:00, 48.36it/s]

Time taken: 0.08 minutes with 30 cores
Done!
Loading BokehJS ...
Loading BokehJS ...
In [19]:
# Parameter Dictionary
#C_range = [1e-6,1e-5,1e-4,5e-4,1e-3,5e-3,1e-2,1e-1]
C_range = [0.0001, 0.0002, 0.0003, 0.0004, 0.0005, 0.0006, 0.0007, 0.0008, 0.0009, 0.001]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                                 

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')   

Running ...

1/2: 100%|███████████████████████████████████████████████████████████████████████████| 10/10 [00:00<00:00, 14634.70it/s]
2/2: 100%|████████████████████████████████████████████████████████████████████████████| 100/100 [00:02<00:00, 43.82it/s]

Time taken: 0.11 minutes with 30 cores
Done!
Loading BokehJS ...
Loading BokehJS ...
In [20]:
# Build Model
model = cb.model.SVM(C=0.0005, kernel="linear")     
YPredTrain = model.train(XTrainKnn, YTrain)
YPredTest = model.test(XTestKnn)

# Put YTrain and YPredTrain in a List
EvalTrain = [YTrain, YPredTrain]

# Put YTest and YPrestTest in a List
EvalTest = [YTest, YPredTest]

# Evaluate Model (include Test Dataset)
model.evaluate(testset=EvalTest)
Loading BokehJS ...
In [21]:
# Extract X Data
XBoot = DataTable2[PeakList]
XBootLog = np.log(XBoot)
XBootScale = cb.utils.scale(XBootLog, method='auto')
#XBootKnn = XBootScale #cb.utils.knnimpute(XBootScale, k=3)
YPredBoot = model.train(XBootScale, Y)

# Build Boostrap Models
bootmodel = cb.bootstrap.Per(model, bootnum=100) 
bootmodel.run()

# Boostrap Evaluate Model (include Test Dataset)
bootmodel.evaluate(trainset=EvalTrain, testset=EvalTest)  

Number of cores set to: 128

100%|█████████████████████████████████████████████████████████████████████████████████| 100/100 [00:01<00:00, 70.31it/s]

Time taken: 0.18 minutes with 128 cores
Loading BokehJS ...
In [22]:
coef = model.model.coef_ # extract coefficient
important_feat = abs(coef[0])
#get indices of those important features
idx = important_feat.argsort(kind= "quicksort")
idx= idx[::-1][:5]
#get more information from peaktable:
top_met = PeakList.iloc[idx]
top_met_info = PeakTable.iloc[top_met.index-1]
#5 most important metabolites
top_met_info
Out[22]:
Idx Name Label Perc_missing
41 41 M41 1-linoleoyl-GPA (18:2)* 0.303214
618 618 M618 lysine 0.000000
123 123 M123 2-aminoheptanoate 0.000000
476 476 M476 gamma-glutamyl-alpha-lysine 0.060643
576 576 M576 Isobar: fructose 6-phosphate, fructose 1-phosp... 2.546998

ML RF:¶

In [23]:
# Parameter Dictionary
depth = list(range(1,11))
leaf_asfraction = [0.01,0.05,0.1,0.15,0.2,0.25,0.3,0.35,0.4,0.45,0.5]

param_dict = dict(max_depth=depth,
                  min_samples_leaf=leaf_asfraction,
                  max_features='sqrt',
                  criterion='gini',
                  min_samples_split=2,
                  max_leaf_nodes=None,
                  n_estimators=100)

# Initialise
cv = cb.cross_val.KFold(model=cb.model.RF,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)       
                              

# Run and Plot
cv.run()  
cv.plot(metric='auc', color_beta=[5,5,3])
cv.plot(metric='r2q2', color_beta=[5,5,3])

Running ...

1/2:   0%|                                                                                      | 0/110 [00:00<?, ?it/s]Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
1/2: 100%|████████████████████████████████████████████████████████████████████████████| 110/110 [01:05<00:00,  1.67it/s]
2/2: 100%|██████████████████████████████████████████████████████████████████████████| 1100/1100 [00:52<00:00, 20.86it/s]

Time taken: 2.07 minutes with 30 cores
Done!

/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
Loading BokehJS ... 
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
Loading BokehJS ...
In [24]:
# Build Model
model = cb.model.RF(max_depth=7,
                    min_samples_leaf=0.05,
                    max_features='sqrt',
                    criterion='gini',
                    min_samples_split=2,
                    max_leaf_nodes=None,
                    n_estimators=100) 
YPredTrain = model.train(XTrainKnn, YTrain)
YPredTest = model.test(XTestKnn)

# Put YTrain and YPredTrain in a List
EvalTrain = [YTrain, YPredTrain]

# Put YTest and YPrestTest in a List
EvalTest = [YTest, YPredTest]

# Evaluate Model (include Test Dataset
model.evaluate(testset=EvalTest)
Loading BokehJS ...
In [25]:
# Extract X Data
XBoot = DataTable2[PeakList]
XBootLog = np.log(XBoot)
XBootScale = cb.utils.scale(XBootLog, method='auto')
#XBootKnn = XBootScale #cb.utils.knnimpute(XBootScale, k=3)
YPredBoot = model.train(XBootScale, Y)

# Build Boostrap Models
bootmodel = cb.bootstrap.Per(model, bootnum=100) 
bootmodel.run()

# Boostrap Evaluate Model (include Test Dataset)
bootmodel.evaluate(trainset=EvalTrain, testset=EvalTest)   

Number of cores set to: 128

100%|█████████████████████████████████████████████████████████████████████████████████| 100/100 [00:01<00:00, 64.56it/s]

Time taken: 0.22 minutes with 128 cores
Loading BokehJS ...
In [26]:
# for important metabolies / features:
important_feat = model.model.feature_importances_
important_feat = abs(important_feat)
#get indices of those important features
idx = important_feat.argsort(kind= "quicksort")
idx= idx[::-1][:5]
#get more information from peaktable:
top_met = PeakList.iloc[idx]
top_met_info = PeakTable.iloc[top_met.index-1]
#5 most important metabolites
top_met_info
Out[26]:
Idx Name Label Perc_missing
41 41 M41 1-linoleoyl-GPA (18:2)* 0.303214
742 742 M742 oleoyl-arachidonoyl-glycerol (18:1/20:4) [2]* 0.363857
476 476 M476 gamma-glutamyl-alpha-lysine 0.060643
123 123 M123 2-aminoheptanoate 0.000000
143 143 M143 2-linoleoylglycerol (18:2) 0.545785

50% removed: median imputation¶

In [27]:
# impute the missing values and create the missing value indicator variables for each numeric column.
df_numeric = miss_50_DataTable._get_numeric_data().copy()
numeric_cols = df_numeric.columns.values
# imputation for numeric values:
for col in numeric_cols:
    df_numeric[col].fillna(df_numeric[col].median(), inplace=True)

/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/numpy/lib/nanfunctions.py:1117: RuntimeWarning: Mean of empty slice
  return np.nanmean(a, axis, out=out, keepdims=keepdims)
In [28]:
#check if anything nans remaining due to only nans in data, remove them
only_missings= df_numeric.columns[df_numeric.isna().any()].tolist()# remove them because of too many NANs in column.
df_numeric= df_numeric.drop(only_missings, axis=1)
In [29]:
#for non-numeric columns using mode:
df_non_numeric = miss_50_DataTable.select_dtypes(exclude=np.number).copy()
non_numeric_cols = df_non_numeric.columns.values

for col in non_numeric_cols:
    df_non_numeric[col].fillna(df_non_numeric[col].mode()[0], inplace=True)
    
#concatinate the non and numeric df:
imputed_50_miss = pd.concat([df_non_numeric,df_numeric], axis=1, join='inner')
#plot the missing vs no missing after imputation:
import matplotlib.pyplot as plt
cols = imputed_50_miss.columns
colors = ['#008000', '#ff0000']
m_heatmap= sns.heatmap(imputed_50_miss[cols].isnull(), cmap=sns.color_palette(['#1f77b4', '#ff0000']),
            cbar=False, linecolor='gray')
m_heatmap.set_xlabel('Features')
m_heatmap.set_title('Heatmap after median imputation for 50% removed data')
m_heatmap.legend(title='missing: red, present: blue', labels=['Present', 'Missing'], loc='upper right')
Out[29]:
<matplotlib.legend.Legend at 0x7f59e1aad8d0>
In [30]:
imputed_50_miss.columns[imputed_50_miss.isna().any()].tolist()
Out[30]:
[]
In [31]:
# Clean PeakTable and Extract PeakList
PercMiss = PeakTable['Perc_missing']  
PeakTableClean = PeakTable[(PercMiss < 20)] 
PeakList = PeakTableClean['Name']  

# Select Subset of Data
DataTable2 = imputed_50_miss[(imputed_50_miss.Class == 1) | (imputed_50_miss.Class == 0)]

# Create a Binary Y Vector 
Outcomes = DataTable2['Class']
Y = Outcomes.values 

# Split Data into Train (2/3) and Test (1/3)
DataTrain, DataTest, YTrain, YTest = train_test_split(DataTable2, Y, test_size=1/3, stratify=Y, random_state=85)

# Extract Train Data 
XTrain = DataTrain[PeakList]                                    
XTrainLog = np.log(XTrain)                                          
XTrainScale, mu, sigma = cb.utils.scale(XTrainLog, method='auto', return_mu_sigma=True)              
XTrainKnn = XTrainScale# cb.utils.knnimpute(XTrainScale, k=3)    

# Extract Test Data
XTest = DataTest[PeakList]                                     
XTestLog = np.log(XTest)                                          
XTestScale = cb.utils.scale(XTestLog, method='auto', mu=mu, sigma=sigma) 
XTestKnn = XTestScale #cb.utils.knnimpute(XTestScale, k=3)
In [32]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...

1/2:   0%|                                                                                        | 0/7 [00:00<?, ?it/s]Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
1/2: 100%|████████████████████████████████████████████████████████████████████████████████| 7/7 [00:42<00:00,  6.03s/it]
2/2: 100%|█████████████████████████████████████████████████████████████████████████████| 70/70 [00:00<00:00, 190.28it/s]

Time taken: 1.64 minutes with 30 cores
Done!
Loading BokehJS ...
Loading BokehJS ...
In [33]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...

1/2: 100%|██████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 8340.95it/s]
2/2: 100%|█████████████████████████████████████████████████████████████████████████████| 70/70 [00:00<00:00, 159.92it/s]

Time taken: 0.03 minutes with 30 cores
Done!
Loading BokehJS ...
Loading BokehJS ...
In [34]:
# Parameter Dictionary
#C_range = [1e-6,1e-5,1e-4,5e-4,1e-3,5e-3,1e-2,1e-1]
C_range = [0.0001, 0.0002, 0.0003, 0.0004, 0.0005, 0.0006, 0.0007, 0.0008, 0.0009, 0.001]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                                 

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')   

Running ...

1/2: 100%|███████████████████████████████████████████████████████████████████████████| 10/10 [00:00<00:00, 12572.85it/s]
2/2: 100%|███████████████████████████████████████████████████████████████████████████| 100/100 [00:00<00:00, 113.39it/s]

Time taken: 0.03 minutes with 30 cores
Done!
Loading BokehJS ...
Loading BokehJS ...
In [35]:
# Build Model
model = cb.model.SVM(C=0.0005, kernel="linear")     
YPredTrain = model.train(XTrainKnn, YTrain)
YPredTest = model.test(XTestKnn)

# Put YTrain and YPredTrain in a List
EvalTrain = [YTrain, YPredTrain]

# Put YTest and YPrestTest in a List
EvalTest = [YTest, YPredTest]

# Evaluate Model (include Test Dataset)
model.evaluate(testset=EvalTest)
Loading BokehJS ...
In [36]:
# Extract X Data
XBoot = DataTable2[PeakList]
XBootLog = np.log(XBoot)
XBootScale = cb.utils.scale(XBootLog, method='auto')
#XBootKnn = XBootScale #cb.utils.knnimpute(XBootScale, k=3)
YPredBoot = model.train(XBootScale, Y)

# Build Boostrap Models
bootmodel = cb.bootstrap.Per(model, bootnum=100) 
bootmodel.run()

# Boostrap Evaluate Model (include Test Dataset)
bootmodel.evaluate(trainset=EvalTrain, testset=EvalTest)  

Number of cores set to: 128

100%|█████████████████████████████████████████████████████████████████████████████████| 100/100 [00:01<00:00, 56.26it/s]

Time taken: 0.11 minutes with 128 cores
Loading BokehJS ...
In [37]:
coef = model.model.coef_ # extract coefficient
important_feat = abs(coef[0])
#get indices of those important features
idx = important_feat.argsort(kind= "quicksort")
idx= idx[::-1][:5]
#get more information from peaktable:
top_met = PeakList.iloc[idx]
top_met_info = PeakTable.iloc[top_met.index-1]
#5 most important metabolites
top_met_info
Out[37]:
Idx Name Label Perc_missing
907 907 M907 theophylline 14.918132
910 910 M910 threonine 0.000000
305 305 M305 androstenediol (3beta,17beta) disulfate (2) 0.000000
39 39 M39 1-linoleoyl-2-arachidonoyl-GPC (18:2/20:4n6)* 1.030928
122 122 M122 2-aminobutyrate 0.000000

RF¶

In [38]:
# Parameter Dictionary
depth = list(range(1,11))
leaf_asfraction = [0.01,0.05,0.1,0.15,0.2,0.25,0.3,0.35,0.4,0.45,0.5]

param_dict = dict(max_depth=depth,
                  min_samples_leaf=leaf_asfraction,
                  max_features='sqrt',
                  criterion='gini',
                  min_samples_split=2,
                  max_leaf_nodes=None,
                  n_estimators=100)

# Initialise
cv = cb.cross_val.KFold(model=cb.model.RF,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)       
                              

# Run and Plot
cv.run()  
cv.plot(metric='auc', color_beta=[5,5,3])
cv.plot(metric='r2q2', color_beta=[5,5,3])

Running ...

1/2:   0%|                                                                                      | 0/110 [00:00<?, ?it/s]Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
1/2: 100%|████████████████████████████████████████████████████████████████████████████| 110/110 [01:44<00:00,  1.05it/s]
2/2: 100%|██████████████████████████████████████████████████████████████████████████| 1100/1100 [00:43<00:00, 25.17it/s]

Time taken: 2.54 minutes with 30 cores
Done!

/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
Loading BokehJS ... 
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/utils/color_scale.py:9: FutureWarning: Support for multi-dimensional indexing (e.g. `obj[:, None]`) is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.
  x_init = scaler.fit_transform(x[:, np.newaxis]).flatten()
Loading BokehJS ...
In [39]:
# Build Model
model = cb.model.RF(max_depth=7,
                    min_samples_leaf=0.05,
                    max_features='sqrt',
                    criterion='gini',
                    min_samples_split=2,
                    max_leaf_nodes=None,
                    n_estimators=100) 
YPredTrain = model.train(XTrainKnn, YTrain)
YPredTest = model.test(XTestKnn)

# Put YTrain and YPredTrain in a List
EvalTrain = [YTrain, YPredTrain]

# Put YTest and YPrestTest in a List
EvalTest = [YTest, YPredTest]

# Evaluate Model (include Test Dataset
model.evaluate(testset=EvalTest)
Loading BokehJS ...
In [40]:
# Extract X Data
XBoot = DataTable2[PeakList]
XBootLog = np.log(XBoot)
XBootScale = cb.utils.scale(XBootLog, method='auto')
XBootKnn = XBootScale #cb.utils.knnimpute(XBootScale, k=3)
YPredBoot = model.train(XBootKnn, Y)

# Build Boostrap Models
bootmodel = cb.bootstrap.Per(model, bootnum=100) 
bootmodel.run()

# Boostrap Evaluate Model (include Test Dataset)
bootmodel.evaluate(trainset=EvalTrain, testset=EvalTest)   

Number of cores set to: 128

100%|█████████████████████████████████████████████████████████████████████████████████| 100/100 [00:01<00:00, 58.88it/s]

Time taken: 0.16 minutes with 128 cores
Loading BokehJS ...
In [41]:
# for important metabolies / features:
important_feat = model.model.feature_importances_
important_feat = abs(important_feat)
#get indices of those important features
idx = important_feat.argsort(kind= "quicksort")
idx= idx[::-1][:5]
#get more information from peaktable:
top_met = PeakList.iloc[idx]
top_met_info = PeakTable.iloc[top_met.index-1]
#5 most important metabolites
top_met_info
Out[41]:
Idx Name Label Perc_missing
86 86 M86 1-stearoyl-2-arachidonoyl-GPI (18:0/20:4) 0.060643
29 29 M29 1-arachidonoyl-GPC (20:4n6)* 0.000000
613 613 M613 linoleoyl-linoleoyl-glycerol (18:2/18:2) [1]* 0.000000
221 221 M221 4-guanidinobutanoate 0.000000
77 77 M77 1-palmitoyl-GPA (16:0) 2.910855 
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
Using Theano backend.
In [ ]:

removing irrelevant data for 10% changes:¶

In [42]:
#irr_rem_change_10_DataSet= change_50_DataSet.drop(['M464'], axis=1)
irr_rem_change_10_DataSet= change_10_DataSet.dropna()
irr_rem_change_10_DataSet
Out[42]:
Class Hormone M1 M2 M3 M4 M5 M6 M7 M8 ... M940 M941 M942 M943 M944 M945 M946 M947 M948 M949
SampleID

0 rows × 951 columns

removing irrelevant data for 10% missing:¶

In [43]:
miss_10_DataSet.dropna()
Out[43]:
Class Hormone M1 M2 M3 M4 M5 M6 M7 M8 ... M940 M941 M942 M943 M944 M945 M946 M947 M948 M949
SampleID

0 rows × 951 columns

In [44]:
#########################################

for ML dirty dataset:¶

dirty 10% missing:¶

In [16]:
# Clean PeakTable and Extract PeakList
PercMiss = PeakTable['Perc_missing']  
PeakTableClean = PeakTable[(PercMiss < 20)] 
PeakList = PeakTableClean['Name']  

# Select Subset of Data
DataTable2 = miss_10_DataSet[(miss_10_DataSet.Class == 1) | (miss_10_DataSet.Class == 0)]

# Create a Binary Y Vector 
Outcomes = DataTable2['Class']
Y = Outcomes.values 

# Split Data into Train (2/3) and Test (1/3)
DataTrain, DataTest, YTrain, YTest = train_test_split(DataTable2, Y, test_size=1/3, stratify=Y, random_state=85)

# Extract Train Data 
XTrain = DataTrain[PeakList]                                    
XTrainLog = np.log(XTrain)                                          
XTrainScale, mu, sigma = cb.utils.scale(XTrainLog, method='auto', return_mu_sigma=True)              
XTrainKnn = XTrainScale#cb.utils.knnimpute(XTrainScale, k=3)    

# Extract Test Data
XTest = DataTest[PeakList]                                     
XTestLog = np.log(XTest)                                          
XTestScale = cb.utils.scale(XTestLog, method='auto', mu=mu, sigma=sigma) 
XTestKnn = XTestScale# cb.utils.knnimpute(XTestScale, k=3)
In [17]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...

1/2: 100%|███████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 817.40it/s]

TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.

1/2:   0%|                                                                                        | 0/7 [00:00<?, ?it/s]
---------------------------------------------------------------------------
_RemoteTraceback                          Traceback (most recent call last)
_RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 428, in _process_worker
    r = call_item()
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 275, in __call__
    return self.fn(*self.args, **self.kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py", line 620, in __call__
    return self.func(*args, **kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in __call__
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in <listcomp>
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py", line 113, in _calc_full_loop
    model_i.train(self.X, self.Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py", line 48, in train
    self.model.fit(X, Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py", line 196, in fit
    accept_large_sparse=False,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py", line 581, in _validate_data
    X, y = check_X_y(X, y, **check_params)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 976, in check_X_y
    estimator=estimator,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 800, in check_array
    _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 116, in _assert_all_finite
    type_err, msg_dtype if msg_dtype is not None else X.dtype
ValueError: Input contains NaN, infinity or a value too large for dtype('float64').
"""

The above exception was the direct cause of the following exception:

ValueError                                Traceback (most recent call last)
~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     61         try:
---> 62             full = Parallel(n_jobs=self.n_cores)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     63         except:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1097             with self._backend.retrieval_context():
-> 1098                 self.retrieve()
   1099             # Make sure that we get a last message telling us we are done

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in retrieve(self)
    974                 if getattr(self._backend, 'supports_timeout', False):
--> 975                     self._output.extend(job.get(timeout=self.timeout))
    976                 else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in wrap_future_result(future, timeout)
    566         try:
--> 567             return future.result(timeout=timeout)
    568         except CfTimeoutError as e:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in result(self, timeout)
    431             elif self._state == FINISHED:
--> 432                 return self.__get_result()
    433             else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in __get_result(self)
    383         if self._exception:
--> 384             raise self._exception
    385         else:

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

During handling of the above exception, another exception occurred:

ValueError                                Traceback (most recent call last)
/tmp/ipykernel_519792/704266850.py in <module>
     13 
     14 # Run and Plot
---> 15 cv.run()
     16 cv.plot(metric='auc')
     17 cv.plot(metric='r2q2')

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/BaseCrossVal.py in run(self)
    100             print("returning stats at 'x' epoch interval during training until epoch={}.".format(epoch_list[-1]))
    101         else:
--> 102             self.calc_ypred()
    103         self.calc_stats()
    104         print("Done!")

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     63         except:
     64             print("TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.")
---> 65             full = Parallel(n_jobs=1)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     66         self.ypred_full = []
     67         self.x_scores_full = []

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1083             # remaining jobs.
   1084             self._iterating = False
-> 1085             if self.dispatch_one_batch(iterator):
   1086                 self._iterating = self._original_iterator is not None
   1087 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in dispatch_one_batch(self, iterator)
    899                 return False
    900             else:
--> 901                 self._dispatch(tasks)
    902                 return True
    903 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in _dispatch(self, batch)
    817         with self._lock:
    818             job_idx = len(self._jobs)
--> 819             job = self._backend.apply_async(batch, callback=cb)
    820             # A job can complete so quickly than its callback is
    821             # called before we get here, causing self._jobs to

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in apply_async(self, func, callback)
    206     def apply_async(self, func, callback=None):
    207         """Schedule a func to be run"""
--> 208         result = ImmediateResult(func)
    209         if callback:
    210             callback(result)

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in __init__(self, batch)
    595         # Don't delay the application, to avoid keeping the input
    596         # arguments in memory
--> 597         self.results = batch()
    598 
    599     def get(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in <listcomp>(.0)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in _calc_full_loop(self, i)
    111         if model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid" or model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid":
    112             model_i.compiled = False
--> 113         model_i.train(self.X, self.Y)
    114         ypred_full_i = model_i.test(self.X)
    115         ypred_full = ypred_full_i

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py in train(self, X, Y)
     46 
     47         # Fit the model
---> 48         self.model.fit(X, Y)
     49 
     50         # Predict_proba was designed for multi-groups...

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py in fit(self, X, y, sample_weight)
    194                 order="C",
    195                 accept_sparse="csr",
--> 196                 accept_large_sparse=False,
    197             )
    198 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py in _validate_data(self, X, y, reset, validate_separately, **check_params)
    579                 y = check_array(y, **check_y_params)
    580             else:
--> 581                 X, y = check_X_y(X, y, **check_params)
    582             out = X, y
    583 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_X_y(X, y, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, multi_output, ensure_min_samples, ensure_min_features, y_numeric, estimator)
    974         ensure_min_samples=ensure_min_samples,
    975         ensure_min_features=ensure_min_features,
--> 976         estimator=estimator,
    977     )
    978 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_array(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, estimator)
    798 
    799         if force_all_finite:
--> 800             _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
    801 
    802     if ensure_min_samples > 0:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in _assert_all_finite(X, allow_nan, msg_dtype)
    114             raise ValueError(
    115                 msg_err.format(
--> 116                     type_err, msg_dtype if msg_dtype is not None else X.dtype
    117                 )
    118             )

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

dirty 50% remove dataset¶

In [18]:
# Clean PeakTable and Extract PeakList
PercMiss = PeakTable['Perc_missing']  
PeakTableClean = PeakTable[(PercMiss < 20)] 
PeakList = PeakTableClean['Name']  

# Select Subset of Data
DataTable2 = miss_50_DataSet[(miss_50_DataSet.Class == 1) | (miss_50_DataSet.Class == 0)]

# Create a Binary Y Vector 
Outcomes = DataTable2['Class']
Y = Outcomes.values 

# Split Data into Train (2/3) and Test (1/3)
DataTrain, DataTest, YTrain, YTest = train_test_split(DataTable2, Y, test_size=1/3, stratify=Y, random_state=85)

# Extract Train Data 
XTrain = DataTrain[PeakList]                                    
XTrainLog = np.log(XTrain)                                          
XTrainScale, mu, sigma = cb.utils.scale(XTrainLog, method='auto', return_mu_sigma=True)              
XTrainKnn = XTrainScale#cb.utils.knnimpute(XTrainScale, k=3)    

# Extract Test Data
XTest = DataTest[PeakList]                                     
XTestLog = np.log(XTest)                                          
XTestScale = cb.utils.scale(XTestLog, method='auto', mu=mu, sigma=sigma) 
XTestKnn = XTestScale# cb.utils.knnimpute(XTestScale, k=3)
In [19]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...


1/2: 100%|██████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 3062.17it/s]

TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.


1/2:   0%|                                                                                        | 0/7 [00:00<?, ?it/s]
---------------------------------------------------------------------------
_RemoteTraceback                          Traceback (most recent call last)
_RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 428, in _process_worker
    r = call_item()
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 275, in __call__
    return self.fn(*self.args, **self.kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py", line 620, in __call__
    return self.func(*args, **kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in __call__
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in <listcomp>
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py", line 113, in _calc_full_loop
    model_i.train(self.X, self.Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py", line 48, in train
    self.model.fit(X, Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py", line 196, in fit
    accept_large_sparse=False,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py", line 581, in _validate_data
    X, y = check_X_y(X, y, **check_params)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 976, in check_X_y
    estimator=estimator,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 800, in check_array
    _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 116, in _assert_all_finite
    type_err, msg_dtype if msg_dtype is not None else X.dtype
ValueError: Input contains NaN, infinity or a value too large for dtype('float64').
"""

The above exception was the direct cause of the following exception:

ValueError                                Traceback (most recent call last)
~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     61         try:
---> 62             full = Parallel(n_jobs=self.n_cores)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     63         except:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1097             with self._backend.retrieval_context():
-> 1098                 self.retrieve()
   1099             # Make sure that we get a last message telling us we are done

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in retrieve(self)
    974                 if getattr(self._backend, 'supports_timeout', False):
--> 975                     self._output.extend(job.get(timeout=self.timeout))
    976                 else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in wrap_future_result(future, timeout)
    566         try:
--> 567             return future.result(timeout=timeout)
    568         except CfTimeoutError as e:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in result(self, timeout)
    431             elif self._state == FINISHED:
--> 432                 return self.__get_result()
    433             else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in __get_result(self)
    383         if self._exception:
--> 384             raise self._exception
    385         else:

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

During handling of the above exception, another exception occurred:

ValueError                                Traceback (most recent call last)
/tmp/ipykernel_519792/704266850.py in <module>
     13 
     14 # Run and Plot
---> 15 cv.run()
     16 cv.plot(metric='auc')
     17 cv.plot(metric='r2q2')

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/BaseCrossVal.py in run(self)
    100             print("returning stats at 'x' epoch interval during training until epoch={}.".format(epoch_list[-1]))
    101         else:
--> 102             self.calc_ypred()
    103         self.calc_stats()
    104         print("Done!")

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     63         except:
     64             print("TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.")
---> 65             full = Parallel(n_jobs=1)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     66         self.ypred_full = []
     67         self.x_scores_full = []

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1083             # remaining jobs.
   1084             self._iterating = False
-> 1085             if self.dispatch_one_batch(iterator):
   1086                 self._iterating = self._original_iterator is not None
   1087 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in dispatch_one_batch(self, iterator)
    899                 return False
    900             else:
--> 901                 self._dispatch(tasks)
    902                 return True
    903 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in _dispatch(self, batch)
    817         with self._lock:
    818             job_idx = len(self._jobs)
--> 819             job = self._backend.apply_async(batch, callback=cb)
    820             # A job can complete so quickly than its callback is
    821             # called before we get here, causing self._jobs to

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in apply_async(self, func, callback)
    206     def apply_async(self, func, callback=None):
    207         """Schedule a func to be run"""
--> 208         result = ImmediateResult(func)
    209         if callback:
    210             callback(result)

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in __init__(self, batch)
    595         # Don't delay the application, to avoid keeping the input
    596         # arguments in memory
--> 597         self.results = batch()
    598 
    599     def get(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in <listcomp>(.0)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in _calc_full_loop(self, i)
    111         if model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid" or model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid":
    112             model_i.compiled = False
--> 113         model_i.train(self.X, self.Y)
    114         ypred_full_i = model_i.test(self.X)
    115         ypred_full = ypred_full_i

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py in train(self, X, Y)
     46 
     47         # Fit the model
---> 48         self.model.fit(X, Y)
     49 
     50         # Predict_proba was designed for multi-groups...

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py in fit(self, X, y, sample_weight)
    194                 order="C",
    195                 accept_sparse="csr",
--> 196                 accept_large_sparse=False,
    197             )
    198 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py in _validate_data(self, X, y, reset, validate_separately, **check_params)
    579                 y = check_array(y, **check_y_params)
    580             else:
--> 581                 X, y = check_X_y(X, y, **check_params)
    582             out = X, y
    583 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_X_y(X, y, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, multi_output, ensure_min_samples, ensure_min_features, y_numeric, estimator)
    974         ensure_min_samples=ensure_min_samples,
    975         ensure_min_features=ensure_min_features,
--> 976         estimator=estimator,
    977     )
    978 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_array(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, estimator)
    798 
    799         if force_all_finite:
--> 800             _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
    801 
    802     if ensure_min_samples > 0:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in _assert_all_finite(X, allow_nan, msg_dtype)
    114             raise ValueError(
    115                 msg_err.format(
--> 116                     type_err, msg_dtype if msg_dtype is not None else X.dtype
    117                 )
    118             )

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

dirty change 10% dataset:¶

In [20]:
# Clean PeakTable and Extract PeakList
PercMiss = PeakTable['Perc_missing']  
PeakTableClean = PeakTable[(PercMiss < 20)] 
PeakList = PeakTableClean['Name']  

# Select Subset of Data
DataTable2 = change_10_DataSet[(change_10_DataSet.Class == 1) | (change_10_DataSet.Class == 0)]

# Create a Binary Y Vector 
Outcomes = DataTable2['Class']
Y = Outcomes.values 

# Split Data into Train (2/3) and Test (1/3)
DataTrain, DataTest, YTrain, YTest = train_test_split(DataTable2, Y, test_size=1/3, stratify=Y, random_state=85)

# Extract Train Data 
XTrain = DataTrain[PeakList]                                    
XTrainLog = np.log(XTrain)                                          
XTrainScale, mu, sigma = cb.utils.scale(XTrainLog, method='auto', return_mu_sigma=True)              
XTrainKnn = XTrainScale#cb.utils.knnimpute(XTrainScale, k=3)    

# Extract Test Data
XTest = DataTest[PeakList]                                     
XTestLog = np.log(XTest)                                          
XTestScale = cb.utils.scale(XTestLog, method='auto', mu=mu, sigma=sigma) 
XTestKnn = XTestScale# cb.utils.knnimpute(XTestScale, k=3)
In [21]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...



1/2: 100%|██████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 3022.46it/s]

TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.



1/2:   0%|                                                                                        | 0/7 [00:00<?, ?it/s]
---------------------------------------------------------------------------
_RemoteTraceback                          Traceback (most recent call last)
_RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 428, in _process_worker
    r = call_item()
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 275, in __call__
    return self.fn(*self.args, **self.kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py", line 620, in __call__
    return self.func(*args, **kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in __call__
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in <listcomp>
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py", line 113, in _calc_full_loop
    model_i.train(self.X, self.Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py", line 48, in train
    self.model.fit(X, Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py", line 196, in fit
    accept_large_sparse=False,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py", line 581, in _validate_data
    X, y = check_X_y(X, y, **check_params)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 976, in check_X_y
    estimator=estimator,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 800, in check_array
    _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 116, in _assert_all_finite
    type_err, msg_dtype if msg_dtype is not None else X.dtype
ValueError: Input contains NaN, infinity or a value too large for dtype('float64').
"""

The above exception was the direct cause of the following exception:

ValueError                                Traceback (most recent call last)
~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     61         try:
---> 62             full = Parallel(n_jobs=self.n_cores)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     63         except:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1097             with self._backend.retrieval_context():
-> 1098                 self.retrieve()
   1099             # Make sure that we get a last message telling us we are done

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in retrieve(self)
    974                 if getattr(self._backend, 'supports_timeout', False):
--> 975                     self._output.extend(job.get(timeout=self.timeout))
    976                 else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in wrap_future_result(future, timeout)
    566         try:
--> 567             return future.result(timeout=timeout)
    568         except CfTimeoutError as e:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in result(self, timeout)
    431             elif self._state == FINISHED:
--> 432                 return self.__get_result()
    433             else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in __get_result(self)
    383         if self._exception:
--> 384             raise self._exception
    385         else:

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

During handling of the above exception, another exception occurred:

ValueError                                Traceback (most recent call last)
/tmp/ipykernel_519792/704266850.py in <module>
     13 
     14 # Run and Plot
---> 15 cv.run()
     16 cv.plot(metric='auc')
     17 cv.plot(metric='r2q2')

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/BaseCrossVal.py in run(self)
    100             print("returning stats at 'x' epoch interval during training until epoch={}.".format(epoch_list[-1]))
    101         else:
--> 102             self.calc_ypred()
    103         self.calc_stats()
    104         print("Done!")

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     63         except:
     64             print("TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.")
---> 65             full = Parallel(n_jobs=1)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     66         self.ypred_full = []
     67         self.x_scores_full = []

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1083             # remaining jobs.
   1084             self._iterating = False
-> 1085             if self.dispatch_one_batch(iterator):
   1086                 self._iterating = self._original_iterator is not None
   1087 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in dispatch_one_batch(self, iterator)
    899                 return False
    900             else:
--> 901                 self._dispatch(tasks)
    902                 return True
    903 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in _dispatch(self, batch)
    817         with self._lock:
    818             job_idx = len(self._jobs)
--> 819             job = self._backend.apply_async(batch, callback=cb)
    820             # A job can complete so quickly than its callback is
    821             # called before we get here, causing self._jobs to

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in apply_async(self, func, callback)
    206     def apply_async(self, func, callback=None):
    207         """Schedule a func to be run"""
--> 208         result = ImmediateResult(func)
    209         if callback:
    210             callback(result)

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in __init__(self, batch)
    595         # Don't delay the application, to avoid keeping the input
    596         # arguments in memory
--> 597         self.results = batch()
    598 
    599     def get(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in <listcomp>(.0)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in _calc_full_loop(self, i)
    111         if model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid" or model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid":
    112             model_i.compiled = False
--> 113         model_i.train(self.X, self.Y)
    114         ypred_full_i = model_i.test(self.X)
    115         ypred_full = ypred_full_i

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py in train(self, X, Y)
     46 
     47         # Fit the model
---> 48         self.model.fit(X, Y)
     49 
     50         # Predict_proba was designed for multi-groups...

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py in fit(self, X, y, sample_weight)
    194                 order="C",
    195                 accept_sparse="csr",
--> 196                 accept_large_sparse=False,
    197             )
    198 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py in _validate_data(self, X, y, reset, validate_separately, **check_params)
    579                 y = check_array(y, **check_y_params)
    580             else:
--> 581                 X, y = check_X_y(X, y, **check_params)
    582             out = X, y
    583 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_X_y(X, y, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, multi_output, ensure_min_samples, ensure_min_features, y_numeric, estimator)
    974         ensure_min_samples=ensure_min_samples,
    975         ensure_min_features=ensure_min_features,
--> 976         estimator=estimator,
    977     )
    978 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_array(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, estimator)
    798 
    799         if force_all_finite:
--> 800             _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
    801 
    802     if ensure_min_samples > 0:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in _assert_all_finite(X, allow_nan, msg_dtype)
    114             raise ValueError(
    115                 msg_err.format(
--> 116                     type_err, msg_dtype if msg_dtype is not None else X.dtype
    117                 )
    118             )

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

dirty change 50% dataset:¶

In [22]:
# Clean PeakTable and Extract PeakList
PercMiss = PeakTable['Perc_missing']  
PeakTableClean = PeakTable[(PercMiss < 20)] 
PeakList = PeakTableClean['Name']  

# Select Subset of Data
DataTable2 = change_50_DataSet[(change_50_DataSet.Class == 1) | (change_50_DataSet.Class == 0)]

# Create a Binary Y Vector 
Outcomes = DataTable2['Class']
Y = Outcomes.values 

# Split Data into Train (2/3) and Test (1/3)
DataTrain, DataTest, YTrain, YTest = train_test_split(DataTable2, Y, test_size=1/3, stratify=Y, random_state=85)

# Extract Train Data 
XTrain = DataTrain[PeakList]                                    
XTrainLog = np.log(XTrain)                                          
XTrainScale, mu, sigma = cb.utils.scale(XTrainLog, method='auto', return_mu_sigma=True)              
XTrainKnn = XTrainScale#cb.utils.knnimpute(XTrainScale, k=3)    

# Extract Test Data
XTest = DataTest[PeakList]                                     
XTestLog = np.log(XTest)                                          
XTestScale = cb.utils.scale(XTestLog, method='auto', mu=mu, sigma=sigma) 
XTestKnn = XTestScale# cb.utils.knnimpute(XTestScale, k=3)
In [23]:
# Parameter Dictionary
C_range = [0.0001, 0.0005, 0.001, 0.005, 0.01, 0.05,  0.1]
param_dict = dict(C=C_range, kernel="linear")               

# Initialise
cv = cb.cross_val.KFold(model=cb.model.SVM,                      
                        X=XTrainKnn,                                 
                        Y=YTrain,                               
                        param_dict=param_dict,                   
                        folds=5,
                        n_mc=10,
                       n_cores=30)                            

# Run and Plot
cv.run()  
cv.plot(metric='auc')
cv.plot(metric='r2q2')

Running ...




1/2: 100%|██████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 1210.08it/s]

TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.




1/2:   0%|                                                                                        | 0/7 [00:00<?, ?it/s]
---------------------------------------------------------------------------
_RemoteTraceback                          Traceback (most recent call last)
_RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 428, in _process_worker
    r = call_item()
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/externals/loky/process_executor.py", line 275, in __call__
    return self.fn(*self.args, **self.kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py", line 620, in __call__
    return self.func(*args, **kwargs)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in __call__
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py", line 289, in <listcomp>
    for func, args, kwargs in self.items]
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py", line 113, in _calc_full_loop
    model_i.train(self.X, self.Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py", line 48, in train
    self.model.fit(X, Y)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py", line 196, in fit
    accept_large_sparse=False,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py", line 581, in _validate_data
    X, y = check_X_y(X, y, **check_params)
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 976, in check_X_y
    estimator=estimator,
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 800, in check_array
    _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
  File "/home/anepal/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py", line 116, in _assert_all_finite
    type_err, msg_dtype if msg_dtype is not None else X.dtype
ValueError: Input contains NaN, infinity or a value too large for dtype('float64').
"""

The above exception was the direct cause of the following exception:

ValueError                                Traceback (most recent call last)
~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     61         try:
---> 62             full = Parallel(n_jobs=self.n_cores)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     63         except:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1097             with self._backend.retrieval_context():
-> 1098                 self.retrieve()
   1099             # Make sure that we get a last message telling us we are done

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in retrieve(self)
    974                 if getattr(self._backend, 'supports_timeout', False):
--> 975                     self._output.extend(job.get(timeout=self.timeout))
    976                 else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in wrap_future_result(future, timeout)
    566         try:
--> 567             return future.result(timeout=timeout)
    568         except CfTimeoutError as e:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in result(self, timeout)
    431             elif self._state == FINISHED:
--> 432                 return self.__get_result()
    433             else:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/concurrent/futures/_base.py in __get_result(self)
    383         if self._exception:
--> 384             raise self._exception
    385         else:

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

During handling of the above exception, another exception occurred:

ValueError                                Traceback (most recent call last)
/tmp/ipykernel_519792/704266850.py in <module>
     13 
     14 # Run and Plot
---> 15 cv.run()
     16 cv.plot(metric='auc')
     17 cv.plot(metric='r2q2')

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/BaseCrossVal.py in run(self)
    100             print("returning stats at 'x' epoch interval during training until epoch={}.".format(epoch_list[-1]))
    101         else:
--> 102             self.calc_ypred()
    103         self.calc_stats()
    104         print("Done!")

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in calc_ypred(self)
     63         except:
     64             print("TerminatedWorkerError was raised due to excessive memory usage. n_cores was reduced to 1.")
---> 65             full = Parallel(n_jobs=1)(delayed(self._calc_full_loop)(i) for i in tqdm(range(len(self.param_list)), desc="1/2"))
     66         self.ypred_full = []
     67         self.x_scores_full = []

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self, iterable)
   1083             # remaining jobs.
   1084             self._iterating = False
-> 1085             if self.dispatch_one_batch(iterator):
   1086                 self._iterating = self._original_iterator is not None
   1087 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in dispatch_one_batch(self, iterator)
    899                 return False
    900             else:
--> 901                 self._dispatch(tasks)
    902                 return True
    903 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in _dispatch(self, batch)
    817         with self._lock:
    818             job_idx = len(self._jobs)
--> 819             job = self._backend.apply_async(batch, callback=cb)
    820             # A job can complete so quickly than its callback is
    821             # called before we get here, causing self._jobs to

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in apply_async(self, func, callback)
    206     def apply_async(self, func, callback=None):
    207         """Schedule a func to be run"""
--> 208         result = ImmediateResult(func)
    209         if callback:
    210             callback(result)

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/_parallel_backends.py in __init__(self, batch)
    595         # Don't delay the application, to avoid keeping the input
    596         # arguments in memory
--> 597         self.results = batch()
    598 
    599     def get(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in __call__(self)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/joblib/parallel.py in <listcomp>(.0)
    287         with parallel_backend(self._backend, n_jobs=self._n_jobs):
    288             return [func(*args, **kwargs)
--> 289                     for func, args, kwargs in self.items]
    290 
    291     def __reduce__(self):

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/cross_val/KFold.py in _calc_full_loop(self, i)
    111         if model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid" or model_i.__name__ == "cimcb.model.NN_SigmoidSigmoid":
    112             model_i.compiled = False
--> 113         model_i.train(self.X, self.Y)
    114         ypred_full_i = model_i.test(self.X)
    115         ypred_full = ypred_full_i

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py in train(self, X, Y)
     46 
     47         # Fit the model
---> 48         self.model.fit(X, Y)
     49 
     50         # Predict_proba was designed for multi-groups...

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py in fit(self, X, y, sample_weight)
    194                 order="C",
    195                 accept_sparse="csr",
--> 196                 accept_large_sparse=False,
    197             )
    198 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py in _validate_data(self, X, y, reset, validate_separately, **check_params)
    579                 y = check_array(y, **check_y_params)
    580             else:
--> 581                 X, y = check_X_y(X, y, **check_params)
    582             out = X, y
    583 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_X_y(X, y, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, multi_output, ensure_min_samples, ensure_min_features, y_numeric, estimator)
    974         ensure_min_samples=ensure_min_samples,
    975         ensure_min_features=ensure_min_features,
--> 976         estimator=estimator,
    977     )
    978 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_array(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, estimator)
    798 
    799         if force_all_finite:
--> 800             _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
    801 
    802     if ensure_min_samples > 0:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in _assert_all_finite(X, allow_nan, msg_dtype)
    114             raise ValueError(
    115                 msg_err.format(
--> 116                     type_err, msg_dtype if msg_dtype is not None else X.dtype
    117                 )
    118             )

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').
In [24]:
# Build Model
model = cb.model.SVM(C=0.0005, kernel="linear")     
YPredTrain = model.train(XTrainKnn, YTrain)
YPredTest = model.test(XTestKnn)

# Put YTrain and YPredTrain in a List
EvalTrain = [YTrain, YPredTrain]

# Put YTest and YPrestTest in a List
EvalTest = [YTest, YPredTest]

# Evaluate Model (include Test Dataset)
model.evaluate(testset=EvalTest) 

1/2:   0%|                                                                                        | 0/7 [02:03<?, ?it/s]
1/2:   0%|                                                                                        | 0/7 [01:24<?, ?it/s]

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
/tmp/ipykernel_519792/3108593830.py in <module>
      1 # Build Model
      2 model = cb.model.SVM(C=0.0005, kernel="linear")
----> 3 YPredTrain = model.train(XTrainKnn, YTrain)
      4 YPredTest = model.test(XTestKnn)
      5 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/cimcb/model/SVM.py in train(self, X, Y)
     46 
     47         # Fit the model
---> 48         self.model.fit(X, Y)
     49 
     50         # Predict_proba was designed for multi-groups...

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/svm/_base.py in fit(self, X, y, sample_weight)
    194                 order="C",
    195                 accept_sparse="csr",
--> 196                 accept_large_sparse=False,
    197             )
    198 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/base.py in _validate_data(self, X, y, reset, validate_separately, **check_params)
    579                 y = check_array(y, **check_y_params)
    580             else:
--> 581                 X, y = check_X_y(X, y, **check_params)
    582             out = X, y
    583 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_X_y(X, y, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, multi_output, ensure_min_samples, ensure_min_features, y_numeric, estimator)
    974         ensure_min_samples=ensure_min_samples,
    975         ensure_min_features=ensure_min_features,
--> 976         estimator=estimator,
    977     )
    978 

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in check_array(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, estimator)
    798 
    799         if force_all_finite:
--> 800             _assert_all_finite(array, allow_nan=force_all_finite == "allow-nan")
    801 
    802     if ensure_min_samples > 0:

~/.conda/envs/process_env/envs/MetabComparisonBinaryML/lib/python3.7/site-packages/sklearn/utils/validation.py in _assert_all_finite(X, allow_nan, msg_dtype)
    114             raise ValueError(
    115                 msg_err.format(
--> 116                     type_err, msg_dtype if msg_dtype is not None else X.dtype
    117                 )
    118             )

ValueError: Input contains NaN, infinity or a value too large for dtype('float64').
In [ ]: